Governments and international organizations collectively invest trillions of dollars each year in transport infrastructure — roads, railways, tunnels, bridges. When those investments are appraised, two questions dominate: Will it reduce the cost of moving goods? Will it connect workers to better jobs? Both matter. But they miss something increasingly important.
Our research identifies a third benefit that rarely appears in cost-benefit analyses: transport infrastructure expands the market for services. An engineer travelling to inspect a factory, a technician sent to repair equipment, an accountant or IT specialist who works mostly remotely but still needs occasional face-to-face contact — all of them depend on fast, reliable transport infrastructure to reach clients beyond their immediate city or region. Better infrastructure makes those trips feasible, and in doing so, it opens up markets that would otherwise stay local.
This channel is easy to overlook, partly because it is hard to measure. Goods crossing borders leave customs records. Commuters appear in labor or transport surveys. But a consultant traveling to meet a client in another city usually leaves no comparable statistical trace. As services account for more than half of economic activity in a growing number of developing economies, overlooking this channel can lead policy makers to underestimate the benefits of transport infrastructure.
A bridge that makes an otherwise hidden market visible
In a new working paper, we study the opening of the Øresund Bridge between Malmö, Sweden, and Copenhagen, Denmark, in 2000. Because Malmö and Copenhagen are in different countries, payments for services between them are recorded as international transactions – making them directly observable in a way that similar transactions within the same country normally are not.
The bridge sharply reduced travel times and made daily commuting much easier. We combine firm-level information on 43 types of services with data on commuting, wages, workers, housing, and travel times.
The data also allows us to distinguish two types of economic integration that can involve the same physical journey. A plumber living in Sweden crosses the bridge to work for a Swedish company billing a Danish client generates a service export. The same plumber taking a job with a Danish company represents commuting. But many service exports do not require daily travel: a Swedish consultant, for example, may do most of the work remotely and cross the bridge only occasionally to meet a Danish client. Infrastructure can therefore affect services trade over a much wider geographic area than daily commuting. Infrastructure can therefore affect the two channels very differently
Services responded strongly—and over longer distances than commuting
Five years after the bridge opened, service exports to Denmark by firms close to the bridge had risen by about 172 percent, and service imports had risen by about 232 percent. Commuting from Sweden to Denmark increased more than fivefold. Wages and the working-age population rose in southern Sweden, and house prices near the bridge increased by about 13 percent. By contrast, we find no comparable increase in goods trade.
The geographic patterns are particularly revealing. Daily commuting is highly sensitive to distance: adding time to a journey made twice every working day quickly makes it unattractive. Many services, however, require travel only occasionally—to meet a client, supervise a project, install equipment, or provide on-site support. Firms can therefore serve markets that are too far away for daily commuting.
Our model separates these two channels and finds that their relative importance shifts with distance. Close to the bridge, commuting dominates.
Farther away, access to service markets becomes comparably important — and in a simulated setting without international barriers to services trade, the bridge’s aggregate welfare effect in Sweden increases by about 75 percent, with service-trade costs accounting for the vast majority of that gain. When institutional barriers are low, better physical connectivity can integrate service markets far beyond the daily commuting zone.
